Evaluation

FIGURE 5.17: Results of the tutoring scenario for households (Source: Own Representation)

rate was also quite high with almost 85%. For students however only three quarters would use the service. Again WTA and WTP amounts can be found in figure5.17and5.18and both amounted to 5 to 20 Euros.

The last scenario in the survey was the pet sitting which its result are displayed in figure5.19for the household group and in figure5.20for the student group. Similarly to the ratio of some other scenarios 60% of household participants were identifying with the scenario and 42% wanted to use the service. For the students both values came to almost 75%. WTP for households was between 3 to 15 Euros. WTA for students was 5 to 20 Euros.

FIGURE5.18: Results of the tutoring scenario for students (Source:

Own Representation)

FIGURE 5.19: Results of the pet sitting scenario for households (Source: Own Representation)

FIGURE 5.20: Results of the pet sitting scenario for students (Source: Own Representation)

could be the case due to our method of asking people on the streets to fill out the survey if there were groups of people asked to fill out the questionnaire that are in the same age group. However it should not influence the validity of results since we can separate the age groups if necessary. The ratio of gender with 50% to 50% is reflecting no bias towards any gender of participants. Since the age spectrum of our participants is only reaching up to 65 years the retirement in figure5.3 is quite low. Very few people are also self-employed or jobless. Theoretically the low number of these categories could be attributed to the times and places the survey was conducted. Possibly jobless people are less likely to find in shopping districts and self-employed people have less spare time and therefore spent less time in said shopping districts. Since jobless people are not in the target group of DoBeeDo it does not matter much for testing the service. However for the following results we have to keep in mind that the number of retired and self-employed people is too small to conduct any statistical relevance from them. About one third each of our participants were in a relationship, married and without a partner (i.e. single, divorced or widowed). In figure5.5we can see an expected correlation in the ratio of the relationship status with the number of persons in an household, suggesting that mostly relationships live in 2 people households, married couples as a family in 3-4 people household and single persons alone.

In the student group we have a higher percentage of female participation. This can be explained by a higher percentage of woman studying on the campus in Koblenz.

According to the university’s statistics (Studierende nach Erst- und Neueinschreibung - Studierende und Studienanfänger insgesamt und nach Campi (alle Studierenden ohne Beurlaubte) [pdf]) in winter semester 18/19 about 60% of students are female which is very close to the 63% of our survey. The university also keeps track of bachelor and mas-ter students, however it is not separated by campus (Studierende nach Erst- und Neuein-schreibung - Bachelor- und Masterstudiengänge [Excel]). For the whole university of Koblenz-Landau there are about 27% of master students compared to about 26% that we found. Sadly there were no statistics about the age of students available from the univer-sity but logically our results were in a typical range from 18 to 30.

To prove the validity of DoBeeDo’s Business Model we ran a number of calculations based on the data and results that were collected for the scenarios and introduced in chap-ter4. First of all we were interested in how many percent of participants would actually be interested to use the service that is provided by DoBeeDo unrelated to which kind of scenario they would choose. We got very positive results. Only 7.7% of students where choosing not to do any task, see5.21. The reason for not choosing to do any task could for example be that there was no scenario they were interested in or that they already have a well-paid student job and think they have no benefit from the service. To investigate this reason we looked at whether it were the same people that had a job that choose to not use the service. Also in our data we could see that this was likely the case as 4/5 people in

FIGURE 5.21: Percentage of students who would use the service at least once (Source: Own Representation)

this category had a student job. In the houshold group we achieved 100% in people that would like to use the service in at least one scenario (5.22).

We were also interested in the question whether participants were interested in mul-tiple scenarios or just one. This can help us as an indication of the popularity of the service since later on people would also be inclined to use the application for different tasks and become regular customers. Figures 5.23 and 5.24 show the households and students choosing at least n scenarios. We can see that in both groups participants that wanted to use the service were always interested in multiple scenarios and even about half of them were interested in at least 3 out of 5 scenarios. This shows a really good rate and prognosis for a regular use of the app.

It was also important for us to see whether there exists an indication which scenarios would be good to include in DoBeeDo as possible categories, triggers or just focus points.

For this we again displayed the popularity of scenarios for both groups in one graph (5.25). What is interesting to see from this graph was that for all but one scenario the popularity with students and households is correlating. We believe that a reason for the pet sitting scenario differing is actually a combination of students viewing this task as quite enjoyable and not many households that could identify with the scenario as they do

FIGURE 5.22: Percentage of students who would use the service at least once (Source: Own Representation)

FIGURE 5.23: Percentage of households choosing at least n scenar-ios (Source: Own Representation)

FIGURE 5.24: Percentage of students choosing at least n scenarios (Source: Own Representation)

FIGURE 5.25: Percentage of popularity of scenarios (Source: Own Representation)

not own a dog. In figure5.19this low identification rate in the household group can be seen. The conversion to people who want to use the service is however similar to e.g. the gardening scenario. We believe that the tutoring scenario was most popular for households as this was the only scenario that household owners might not be able to do themselves and for students since it was not connected to physical labour and they felt like they could make use of their intellectual competencies. Moreover it is not to be disregarded that the scenario can be seen as the most traditional student job among them. This can explain why also the scenario was the most popular even though it was also seen as the one with the most hardship (5.43). Other than that a high perceived hardship in car cleaning for students might also explain that this was the least popular scenario to choose for students.

The initial reason why we collected personal information in the survey was that we wanted to see whether this personal features would influence the likeliness of someone using the service. Now we already saw that all participants in the household group wanted to use the service anyway so we were interested whether their features would be correlat-ing with the scenarios they chose. The graphs representcorrelat-ing the age distribution of every scenario in the household group (5.28) were compared to the total age distribution for people that wanted to use the service (5.27). Here we can see that the pet sitting and shopping scenario were of higher popularity in the oldest age group, probably due to de-creased mobility. Which would also explain a small percentage in this group for the car scenario and gardening scenario. For the second oldest group it showed a higher interest in shopping, pet sitting and car cleaning. Possibly in this age group gardening is still done as a recreational activity and the tutoring scenario does not apply to them anymore with their own children or not yet as for grandchildren. Surprisingly the second youngest age

FIGURE 5.26: Average hardship of scenarios (Source: Own Repre-sentation)

FIGURE5.27: Age distribution of households willing to use the ser-vice (Source: Own Representation)

group showed relatively low interest in the gardening scenario, while the youngest one showed the highest interest there, and showed a high interest in the shopping scenario compared to the youngest group, where this expectedly was the least popular one.

Since we only have two age groups for students we can only make some vague as-sumption about the comparison of ratios between figure5.29and figure5.31. But from our collected data it seems like with increasing age and/or education students are less likely to choose ’manual labour’ tasks and are choosing rather service oriented or interest-based.

This also shows in the degree distribution of students for each scenario, where bachelor students were especially less likely to choose the cleaning scenario and more likely to choose the tutoring scenario.

Looking at the gender of the participants we could find that in the household group women are far more likely to use the service than men in almost all scenarios (5.32). In the student group we could not find any significance of gender when it comes to doing a task (5.33).

From the scenario popularity based on the relationship status of household participants (5.34) we can see that relationship distribution varies depending on the scenario. Most noticeably here is the shopping scenario which was most popular with single and married participants. As these groups share no relation in age we think that a common factor in this groups could be how time restricted participants are. Participants without a partner or with a family could be less likely to find the time to go shopping themselves and are therefore willing to seek help.

Starting the survey we had the hypothesis that household participants would be more likely to use the service if they have a higher stress level, which is the reason why we collected the data in our survey. Looking at the results in figure5.35 and5.36we could see that this is actually not necessarily the case. As we found out in chapter 4, only 36,7% percent of participants had a medium to high stress level, i.e. higher or equal than 4. This means that for gardening, shopping and tutoring the likelihood for choosing the service is not significantly higher. For pet sitting the likelihood for choosing the service was actually relatively lower, which we believe is due to the fact that busy people are less likely to get a pet in the first place as it is time consuming to take care of them.

We found a significantly higher likelihood for a stressed participant to choose the car cleaning scenario. Since this scenario is the least popular with students and therefore in general the most undesirable one to do the results could be influenced by a different motivation of household participants. Low stress participants might be able to do the tasks themselves but chose to support students, since low stress would only be possible with sufficient time and monetary resources. Motivated by this they might be conscious to not offer really undesirable tasks. Higher stressed individuals however might be even more striving towards getting rid of undesirable tasks to reduce their stress level.

Of course to evaluate whether or not the Business Model can work we also need to

FIGURE5.28:Agedistributionperscenarioforhouseholds(Source:OwnRepresentation)

FIGURE5.29: Age distribution of students willing to use the service (Source: Own Representation)

compare the prices task givers are willing to pay and task doers are expecting to receive.

For this we first calculated the mean and average WTPs and WTAs for the scenarios (5.37, 5.38). From the median values we can see that the chosen prices are really close if not the same. But including all the available data points in the average, it becomes more obvious what can already be partly seen in the median graph: that students expectations are mostly slightly higher than what households are willing to pay. This is actually not surprising, since, as we mentioned in2.3, it is a common phenomenon in research that WTA is higher then WTP without the influence of supply, demand, reference prices or similar. The closeness of these groups however shows a really good result and to show this we had a more detailed look at the data.

In total figure5.39shows that more than 50% of households had the potential to find a matching student.

From figure5.40to5.44we depicted the number of students that are willing to do the task for a certain amount of money. With an amount at least 10 Euro (which are about 75% of all offers) already more than 23% of students that wanted to do a task would be willing to use the service for almost all scenarios. Since households are supposed to pay a fee in the app, which we estimate around 15%, this would still leave us with 40,7% of all offers. The scenarios that were more unusual here were the tutoring scenario, for which the WTA was higher and the pet sitting scenario, for which it was lower. Like explained before we think the reason for a higher WTA and higher popularity was that students were

FIGURE5.30:Agedistributionperscenarioforstudents(Source:OwnRepresentation)

FIGURE 5.31: Degree distribution students (Source: Own Repre-sentation)

FIGURE 5.32: Gender distribution per scenario for households (Source: Own Representation)

FIGURE5.33: Gender distribution per scenario for students (Source:

Own Representation)

FIGURE5.34: Relationship distribution per scenario for households (Source: Own Representation)

FIGURE5.35: Scenarios chosen by households with low stress level (Source: Own Representation)

FIGURE5.36: Scenarios chosen by households with high stress level (Source: Own Representation)

FIGURE 5.37: Mean WTPs and WTAs (Source: Own Representa-tion)

FIGURE5.38: Median WTPs and WTAs (Source: Own Representa-tion)

FIGURE 5.39: Percentage of Households finding a match (Source:

Own Representation)

viewing this task as more "intellectual" work. Also both groups already had an idea of reference prices which not only raised the WTA but also the WTP5.38. We could not find a reason that supported the low WTA for the pet scenario. Students were not choosing a lower price because they perceived it to be an easier task (5.43) also it was not reflected to be more enjoyable in the popularity of the task (5.20). To be discussed is whether a correlation could show if the motivation for a task would be evaluated using a scale.

Lastly we found that the median Willingness to Accept was lower for bachelor stu-dents than for master stustu-dents as shown in figure5.45.

FIGURE5.40: WTP and WTA for gardening scenario (Source: Own Representation)

FIGURE5.41: WTP and WTA for shopping scenario (Source: Own Representation)

FIGURE 5.42: WTP and WTA for car cleaning scenario (Source:

Own Representation)

FIGURE 5.43: WTP and WTA for tutoring scenario (Source: Own Representation)

FIGURE5.44: WTP and WTA for pet sitting scenario (Source: Own Representation)

FIGURE 5.45: Median WTA for master and bachelor students (Source: Own Representation)

Chapter 6